CN110213149B - Virtual network mapping method based on node entropy - Google Patents

Virtual network mapping method based on node entropy Download PDF

Info

Publication number
CN110213149B
CN110213149B CN201910464796.9A CN201910464796A CN110213149B CN 110213149 B CN110213149 B CN 110213149B CN 201910464796 A CN201910464796 A CN 201910464796A CN 110213149 B CN110213149 B CN 110213149B
Authority
CN
China
Prior art keywords
node
virtual
mapping
physical
network
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201910464796.9A
Other languages
Chinese (zh)
Other versions
CN110213149A (en
Inventor
赵季红
颜皓靓
曲桦
赵建龙
潘峰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian University of Posts and Telecommunications
Original Assignee
Xian University of Posts and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xian University of Posts and Telecommunications filed Critical Xian University of Posts and Telecommunications
Priority to CN201910464796.9A priority Critical patent/CN110213149B/en
Publication of CN110213149A publication Critical patent/CN110213149A/en
Application granted granted Critical
Publication of CN110213149B publication Critical patent/CN110213149B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
    • H04L12/46—Interconnection of networks
    • H04L12/4641—Virtual LANs, VLANs, e.g. virtual private networks [VPN]

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention provides a virtual network mapping method based on node entropy, which is characterized in that node resource attributes are considered, and topological attributes such as node degree and clustering coefficient information are added. Firstly, a concept of node entropy is provided, and a resource attribute and a topology attribute of a network are combined to serve as an evaluation standard for measuring the importance of the nodes; secondly, mapping the virtual nodes to physical nodes with the most similar node entropy; and finally, mapping the virtual link to the physical link by adopting a Dijkstra shortest path algorithm. The algorithm greatly improves the acceptance rate of the virtual network, the network profit-to-cost ratio and the resource utilization rate of the underlying network.

Description

Virtual network mapping method based on node entropy
Technical Field
The invention relates to a virtual network, in particular to a virtual network mapping method based on node entropy.
Background
The internet is one of the most important infrastructures of the current human society, people all over the world are closely connected, the life of people is greatly changed from the birth of the internet to the present decades, and particularly, along with the development of computer technology in the last decade, the internet is continuously popularized, so that great convenience is brought to the life of people. A large number of applications and a wide variety of networking technologies are currently running on the internet. An ubiquitous, worldwide, intercommunicating network has been formed, which is an important infrastructure and productivity of today's society, far beyond what was envisioned when it was first set up.
The initial design of the internet was to ensure survivability of the network in special situations, rather than support large-scale communications, and therefore the initial internet architecture design followed the principle of simple access. However, with the continuous popularization and development of networks, the scale of networks is continuously enlarged due to the access of more and more users and devices, and the current network structure becomes more and more complex due to the endless emergence of various new applications, so that the development of networks encounters a new bottleneck. For example, the network cannot be controlled and managed with fine granularity, the service quality of the network cannot be guaranteed, and the support for mobility is insufficient. It can be said that the development of the current network is in a rigid state, new network services and network functions cannot be deployed quickly, and the whole network architecture and core protocol evolve slowly. And with the development of technology and the emergence of new applications, higher requirements are also put forward on the network.
In order to deal with various problems faced by the current Network, there is a constant search in both academic and industrial fields, the most representative of which are Network Virtualization (Network Virtualization) and Software defined networking (Software Defining Network). The traditional internet tightly couples control logic and data forwarding logic to network devices, which complicates network control plane management, and also makes it difficult to directly deploy network control-level technology updates to existing networks, and flexibility and scalability are limited. The software defined network controls all devices in the network uniformly through a logic centralized controller, a switch in the network is only responsible for forwarding network data packets, and the network is divided into a control layer and a forwarding layer through the structure. The structure with separated control and forwarding and the control mechanism with centralized logic are convenient for realizing centralized management and configuration of the global network, are beneficial to realizing the programmability, controllability, expandability and flexibility of the network, and can effectively solve the rigidity phenomenon in the traditional network evolution.
The essence of network virtualization is that resources in a physical network are abstracted, distributed and isolated, so that multiple independent and non-influencing virtual networks are finally carried in the same physical network, and the physical network needs to reasonably organize and schedule the resources to distribute the resources to different virtual networks, so that flexible and reliable virtual network services are provided. The basic entity of network virtualization is a virtual network, which is a virtual topology formed by a set of virtual nodes and virtual links. A plurality of logic networks on the same physical network belong to different service providers, infrastructure providers can provide different network topology resources for different virtual networks, and the use and management of different virtual networks are independent and do not influence each other. Through the network virtualization technology, the utilization rate of network resources can be effectively improved, and the method has important significance for solving the network rigidity problem.
SDN, as a new network architecture, is one of many implementations of network virtualization. The project of Clean Slate, which originates at Stanford university sponsored by the US GENI project, was formally proposed by the professor Nick McKeown in 2009. The method mainly adopts a layering idea to separate the control function and the forwarding function of the traditional network node, and uniformly controls all equipment in the network through a controller with a logic set, and a switch in the network only has the function of forwarding network data packets. This mechanism facilitates centralized management and configuration of the global network. Because the whole network can be regarded as a logic programmable entity by the logic centralized control capability of the SDN, the completion of network virtualization on the SDN is very convenient, the advantages of SDN control and forwarding can be fully utilized, the purpose of fully utilizing network resources can be achieved by utilizing the advantages of network virtualization, and the advantage of resource sharing is exerted to the maximum extent.
The basis of network virtualization is a virtual network. A virtual network constitutes a collection of virtual nodes and virtual links in a virtual topology. In order to better realize resource sharing in a network virtualization environment based on SDN, an effective resource allocation mechanism needs to be adopted. Under the condition of meeting a certain constraint condition, remaining available resources in the underlying physical network are reasonably allocated to each virtual network request according to needs, which is one of core problems for realizing network virtualization, namely a virtual network mapping problem. Since the quality of the virtual network mapping algorithm directly determines the number of virtual networks that can be carried by a physical network, it is particularly important to select which virtual network mapping algorithm to use in a network virtualization environment.
Virtual network mapping is considered an important aspect of network virtualization, the main objective of which is to utilize limited physical resources to meet the demands of node resources and link resources in a virtual network. Most of the existing research works only consider local resources of the nodes, and neglect the resource attributes of the neighbor nodes, so that the resource utilization rate of the underlying network is not high.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a virtual network mapping method based on node entropy, which solves the problem of low utilization rate of underlying network resources.
The invention is realized by the following technical scheme:
a virtual network mapping method based on node entropy comprises the following steps:
step 1, defining node entropy: combining the node degree, the clustering coefficient information and the node saturation of the node to obtain the node entropy of the node;
step 2, calculating the node entropies of all the physical nodes and the virtual nodes according to the definition of the node entropies in the step 1;
step 3, mapping the virtual nodes to the physical nodes with the node entropy closest to the virtual nodes;
and 4, after the node mapping is successful, mapping the virtual link to a bottom-layer physical link by adopting a Dijkstra algorithm. Preferably, the specific process of step 1 is as follows:
the node degree of a node is expressed as:
Figure GDA0002842299820000021
parameter δ if node n and physical node m are directly connectednmThe value is 1, otherwise 0;
the clustering coefficient information of the node is represented as:
Figure GDA0002842299820000022
wherein e isnRepresenting a triangle consisting of node n and any two of its neighborsThe number of (2);
combining the node degree of the node with the clustering coefficient information to obtain the formulas (3) - (5):
Figure GDA0002842299820000023
Figure GDA0002842299820000024
Figure GDA0002842299820000031
wherein nbr (n) represents a set of neighbor nodes of node n, fnRepresenting the sum of the node degree of a node and its neighbor node degrees, gnInformation representing a neighbor node;
function utilized by the above equation (5)
Figure GDA0002842299820000032
To fnAnd gnNormalized to dn；
The node saturation of a node is expressed as:
Figure GDA0002842299820000033
k represents the number of links connected with the node n, R (n) represents the residual computing resources of the node n, and R (n, m) represents the residual link bandwidth resources of the links connected between the node n and the neighbor nodes m;
the node entropy of a node is expressed as:
NE(n)＝NSn×dn (7)
preferably, if the underlying physical network has I physical nodes and J virtual nodes requested by one virtual network, step 3 specifically includes:
step 3.1, mapping the virtual node j, judging whether the residual resource of the physical node i is smaller than the request resource of the virtual node j, if the residual resource of the physical node i is smaller than the request resource of the virtual node j, executing step 3.2, and if the residual resource of the physical node i is larger than the request resource of the virtual node j, executing step 3.3;
step 3.2, I is I +1, and whether I is less than I is judged, if I is less than I, the step 2.1 is returned, otherwise, the virtual node mapping fails, and the virtual network mapping task is ended;
and 3.3, judging whether the node entropies of the physical node i and the virtual node J are closest, if so, mapping the virtual node J to the physical node i, if J is J +1, judging whether J is less than J, if J is less than J, returning to the step 2.1, otherwise, successfully mapping the virtual node, and executing the step 4.
Further, in step 3.3, the method for determining whether the node entropies of the physical node i and the virtual node j are closest to each other includes: and calculating the absolute value of the node entropy difference between the virtual node j and each physical node, wherein if the absolute value of the node entropy difference between the physical node i and the virtual node j is the minimum in all the absolute values, the node entropy of the physical node i is closest to that of the virtual node j.
Preferably, in step 4, when the target paths corresponding to the multiple virtual links pass through the same physical link, a suitable target path is first found for the virtual link with a large bandwidth resource requirement, and then the virtual link with a small bandwidth requirement is considered.
Preferably, in step 4, a shortest path algorithm is adopted to find a shortest target path between physical nodes bearing virtual nodes at two ends of each virtual link in the physical network, and perform virtual link mapping.
Preferably, assuming that there are L virtual links requested by one virtual network, step 4 specifically includes:
step 4.1, sequencing all virtual links from large to small according to bandwidth requirements, and putting results into Virtuallnklist;
step 4.2, deleting the physical link which does not meet the bandwidth requirement of the virtual link l for the virtual link l in the Virtuallnklist to obtain a physical link set which meets the bandwidth requirement of the virtual link l; if the set is empty, executing step 4.3, otherwise executing step 4.4;
step 4.3, the virtual link mapping fails, and the virtual network mapping task is finished;
and 4.4, mapping the virtual link L to the shortest physical link in the physical link set, enabling L to be L +1, judging whether L is less than L, if L is less than L, returning to the step 4.2, otherwise, successfully mapping the virtual link, and ending the virtual network mapping task.
Compared with the prior art, the invention has the following beneficial technical effects:
according to the algorithm based on the node entropy, on one hand, the node degree, the clustering coefficient information and the node saturation are combined together, and the topological attribute and the resource attribute of the network are comprehensively considered, so that the situations that the resource surplus of some bottom-layer nodes or bottom-layer links is too low in the mapping process, and other link bandwidth resources and node computing resources are too many are reduced. On the other hand, because the node degree, the clustering coefficient and the node saturation are the topology attributes (neighbor nodes) of the nodes in the network, and the topology attributes are the situation of considering the position, the invention takes the position attributes of the network into consideration, so that the node resources and the link resources of the network are distributed more uniformly, the resources are utilized to the maximum extent, the virtual requests can be accepted as much as possible, and the network benefit is improved. On the other hand, the invention balances the distribution of the bottom node and the link resource, and reduces the probability of the bottleneck node or the link. The algorithm of the invention improves the acceptance rate of the virtual network, the network profit-to-cost ratio and the resource utilization rate of the underlying network.
Furthermore, when the target paths corresponding to multiple virtual links pass through the same physical link, contention occurs between different virtual links. In order to reduce the influence of competition and ensure the success rate of link mapping, a proper target path is firstly found for a virtual link with a large bandwidth resource requirement, and then a virtual link with a small bandwidth requirement is considered, so that the success of the whole mapping process is promoted.
Furthermore, in order to shorten the length of link mapping and reduce the link mapping overhead, the shortest path algorithm is adopted in the invention, and for each virtual link, the shortest path between the physical nodes bearing the virtual nodes at the two ends of the link is searched in the physical network.
Drawings
Fig. 1 is a flowchart of a virtual network node mapping algorithm based on node entropy.
Fig. 2 is a flowchart of a virtual network link mapping algorithm based on node entropy.
Fig. 3 is a comparison graph of the virtual network request acceptance rate under different algorithms.
FIG. 4 is a network profit-to-cost ratio comparison graph under different algorithms.
FIG. 5 is a comparison graph of average resource utilization of nodes under different algorithms.
Fig. 6 is a comparison graph of average resource utilization of links under different algorithms.
Detailed Description
The present invention will now be described in further detail with reference to specific examples, which are intended to be illustrative, but not limiting, of the invention. The method is mainly used for solving the virtual network mapping problem under the current SDN-based network virtualization environment. In the network virtualization technology based on the SDN, a physical network performs allocation of virtual network resources through a special controller called a monitor, and slices the physical SDN network into a plurality of virtual SDN networks. Virtual network mapping in an SDN environment needs to consider two issues, mapping of a data layer network, and deployment of a controller. Virtual network mapping is considered an important aspect of network virtualization, the main objective of which is to utilize limited physical resources to meet the demands of node resources and link resources in a virtual network. Most of the existing research works only consider local resources of the nodes, and neglect the resource attributes of the neighbor nodes, so that the resource utilization rate of the underlying network is not high. Aiming at how to reasonably allocate resources to map the virtual SDN network to the underlying physical SDN network, the invention provides a novel virtual network mapping algorithm taking node entropy as a node sequencing index on the basis of comprehensively considering network connectivity. The core idea is that the node resource attribute is considered, and the topological attributes such as node degree and clustering coefficient information are added. The algorithm of the invention improves the acceptance rate of the virtual network, the network profit-to-cost ratio and the resource utilization rate of the underlying network.
The invention relates to a virtual network mapping method based on node entropy, which comprises the following steps:
step 1, defining a node entropy concept of a node, specifically associating node saturation of the node with node saturation of a neighbor node of the node by combining node degree and clustering coefficient information of the node.
The node degree of a node is expressed as:
Figure GDA0002842299820000041
the degree of nodes may reflect the ability of a node to establish direct links with surrounding nodes. Parameter δ if node n and node m are directly connectednmThe value is 1, otherwise it is 0.
The clustering coefficient information is expressed as:
Figure GDA0002842299820000051
in the above formula enRepresenting the number of triangles made up of node n and any two of its neighbors.
The clustering coefficient information is used for describing the ratio of the neighbor node to other nodes in the network, and can reflect the closeness of the neighbor node. Wherein e isnRepresenting the number of triangles formed by the node n and any two of its neighbor nodes, the clustering coefficient information can only reflect the node connectivity of the neighbor nodes but cannot reflect the scale of the neighbor nodes, compared with the node degree. Based on the method, the invention provides a new node measurement method, and the information of the clustering coefficient of the node is considered while the information of the neighbor node is considered.
Combining the node degree and the clustering coefficient information to obtain the following formulas (3) - (5), which are expressed as:
Figure GDA0002842299820000052
Figure GDA0002842299820000053
Figure GDA0002842299820000054
nbr (n) in the above formula (3) represents a neighbor node set of the node n, fnThe sum of the node degree of the node and the neighbor node degree of the node is represented, and the scale of the node and the neighbor node is reflected.
Due to cnCan only reflect the connectivity of the neighbor nodes and cannot reflect the scale of the neighbor nodes, and fnThe node degree of a node and the node degrees of its neighbor nodes may be reflected. So that c is represented by the above formula (4)nAnd fnAre combined with each other by gnInformation representing neighbor nodes.
Function utilized by the above equation (5)
Figure GDA0002842299820000055
To fnAnd gnNormalized to dnIn order to better reflect the topological properties of the network.
The node saturation of a node is expressed as:
Figure GDA0002842299820000056
nbr (n) represents a set of neighbor nodes of the node n, k represents the number of links connected to the node n, R (n) represents the remaining computational resources of the node n, and R (n, m) represents the remaining link bandwidth resources of the links connecting the node n and its neighbor nodes m (2 times because the link bandwidth resources are shared by two neighboring nodes).
The node entropy of a node is expressed as:
NE(n)＝NSn×dn (7)
NSnrepresenting the node saturation of node n, dnRepresenting utilization functions
Figure GDA0002842299820000061
To fnAnd gnNormalization is performed.
Step 2, calculating the node entropy NE of all physical nodes according to the definition of the node entropys(i) And node entropy NE of virtual nodesV(j)。
And 3, mapping each virtual node to the physical node closest to the node entropy by using the node entropy as a node sorting index. The flow chart is shown in fig. 1.
It is assumed that the bottom layer physical network is composed of I physical nodes and M physical links, J virtual nodes requested by one virtual network are provided, and L virtual links are provided.
Step 3.1, mapping the virtual node j, judging whether the residual resource of the physical node i is smaller than the request resource of the virtual node j, if the residual resource of the physical node i is smaller than the request resource of the virtual node j, executing step 3.2, and if the residual resource of the physical node i is larger than the request resource of the virtual node j, executing step 3.3;
step 3.2, I is I +1, and whether I is less than I is judged, if I is less than I, the step 3.1 is returned, otherwise the virtual node mapping fails, and the task is ended;
and 3.3, judging whether the node entropies of the physical node i and the virtual node J are closest, if so, mapping the virtual node J to the physical node i, if J is J +1, judging whether J is less than J, if J is less than J, returning to the step 3.1, otherwise, successfully mapping the virtual node, and executing the step 4.
The judgment method is to use the absolute value | NE of the node entropy difference between the virtual node and the physical nodeS(i)-NEV(j) And the physical node which minimizes the absolute value is the target physical node. The method comprises the following steps: meterCalculating absolute value | NE of node entropy difference between virtual node j and each physical nodeS(i)-NEV(j) If the absolute value of the node entropy difference between the physical node i and the virtual node j is the smallest among all the absolute values, the node entropy of the physical node i is closest to that of the virtual node j.
And 4, after the node mapping is successful, mapping the virtual link to a bottom-layer physical link by adopting a Dijkstra algorithm. The flow chart is shown in fig. 2.
And realizing link mapping in the virtual network on the basis of a node mapping algorithm. When the target paths corresponding to multiple virtual links pass through the same physical link, competition occurs between different virtual links. In order to reduce the influence of competition and ensure the success rate of link mapping, a proper target path is firstly found for a virtual link with a large bandwidth resource requirement, and then a virtual link with a small bandwidth requirement is considered, so that the success of the whole mapping process is promoted. Meanwhile, in order to shorten the length of link mapping and reduce the link mapping overhead, the invention adopts a shortest path algorithm to search the shortest path between the physical nodes bearing the virtual nodes at two ends of each virtual link in a physical network, and the specific steps are as follows:
step 4.1, reordering the virtual links from large to small according to the bandwidth requirement, and putting the result into Virtuallnklist;
step 4.2, deleting the physical link which does not meet the bandwidth requirement of the virtual link l for the virtual link l in the Virtuallnklist to obtain a physical link set which meets the bandwidth requirement of the virtual link l; if the set is empty, executing step 4.3, otherwise executing step 4.4;
step 4.3, the virtual link mapping fails, and the virtual network mapping task is finished;
and 4.4, mapping the virtual link L to the shortest physical link in the physical link set, enabling L to be L +1, judging whether L is less than L, if L is less than L, returning to the step 4.2, otherwise, successfully mapping the virtual link, and ending the virtual network mapping task.
Simulation example
The method adopts the virtual network request acceptance rate, the network profit-to-cost ratio and the physical network resource utilization rate as the evaluation indexes for verifying the algorithm performance, is named as an NE-VNE algorithm and carries out comparative analysis with an NR-VNE algorithm and an NS-VNE algorithm. And the NR-VNE algorithm takes the node degree, the clustering coefficient information and the product of the local resources of the nodes and the links as the measurement index of the node importance for mapping. The NS-VNE algorithm defines the node saturation degree, the node saturation degree serves as a mapping standard of the node, and the Dijkstra algorithm is used for link mapping.
Fig. 3 is a comparison graph of the virtual network request acceptance rate under different algorithms. From the graph, as time goes on, more and more virtual networks are arranged in the underlying network, the virtual network request acceptance rate adopting other algorithms is reduced rapidly, and the virtual network acceptance rate of the invention is always stable and only slightly reduced. The method is mainly characterized in that the node degree, the clustering coefficient information and the node saturation are combined together by the provided algorithm based on the node entropy, and the topological attribute and the resource attribute of the network are comprehensively considered. Therefore, the situation that the resource remaining amount of some bottom nodes or bottom links is too low, and other link bandwidth resources and node computing resources are too much in the mapping process is reduced. Therefore, by adopting the algorithm provided by the invention, when the next virtual network request in the underlying network arrives, relatively sufficient nodes and link resources are also available in the underlying network for mapping, and the success rate of receiving the virtual network request is increased.
See fig. 4 for a comparison of the network profit-to-cost ratios under different algorithms. The higher the profit, the more virtual resources can be carried in the same physical network. As can be seen from FIG. 3, the network profit of the algorithm provided by the invention is reduced along with the time, but the overall network profit is higher than that of the NR-VNE algorithm and the NS-VNE algorithm. The method is mainly characterized in that the position attribute of the network is taken into account based on the algorithm of the node entropy, so that the node resources and the link resources of the network are distributed relatively and uniformly, the resources are utilized to the maximum extent, virtual requests as much as possible can be received, and the network benefit is improved.
Fig. 5 is a comparison graph of average resource utilization of nodes under different algorithms, and fig. 6 is a comparison graph of average resource utilization of links under different algorithms. The graph can show that more virtual networks exist in the underlying network at the same time as time goes on, and the network load becomes larger. This is mainly because the distribution of the underlying node and link resources is balanced, reducing the probability of bottleneck nodes or links occurring.

Claims (6)

1. A virtual network mapping method based on node entropy is characterized by comprising the following steps:
step 1, defining node entropy: combining the node degree, the clustering coefficient information and the node saturation of the node to obtain the node entropy of the node;
step 2, calculating the node entropies of all the physical nodes and the virtual nodes according to the definition of the node entropies in the step 1;
step 3, mapping the virtual nodes to the physical nodes with the node entropy closest to the virtual nodes;
step 4, after the node mapping is successful, mapping the virtual link to a bottom-layer physical link by adopting a Dijkstra algorithm;
the specific process of the step 1 is as follows:
the node degree of a node is expressed as:
Figure FDA0002842299810000011
parameter δ if node n and physical node m are directly connectednmThe value is 1, otherwise 0;
the clustering coefficient information of the node is represented as:
Figure FDA0002842299810000012
wherein e isnRepresenting the number of triangles made up of node n and any two of its neighbors;
combining the node degree of the node with the clustering coefficient information to obtain the formulas (3) - (5):
Figure FDA0002842299810000013
Figure FDA0002842299810000021
Figure FDA0002842299810000022
wherein nbr (n) represents a set of neighbor nodes of node n, fnRepresenting the sum of the node degree of a node and its neighbor node degrees, gnInformation representing a neighbor node;
function utilized by the above equation (5)
Figure FDA0002842299810000023
To fnAnd gnNormalized to dn；
The node saturation of a node is expressed as:
Figure FDA0002842299810000024
k represents the number of links connected with the node n, R (n) represents the residual computing resources of the node n, and R (n, m) represents the residual link bandwidth resources of the links connected between the node n and the neighbor nodes m;
the node entropy of a node is expressed as:
NE(n)＝NSn×dn (7)。
2. the method for mapping a virtual network based on node entropy according to claim 1, wherein, assuming that there are I physical nodes in the underlying physical network and J virtual nodes requested by one virtual network, step 3 specifically includes:
step 3.1, mapping the virtual node j, judging whether the residual resource of the physical node i is smaller than the request resource of the virtual node j, if the residual resource of the physical node i is smaller than the request resource of the virtual node j, executing step 3.2, and if the residual resource of the physical node i is larger than the request resource of the virtual node j, executing step 3.3;
step 3.2, I is I +1, and whether I is less than I is judged, if I is less than I, the step 3.1 is returned, otherwise, the virtual node mapping fails, and the virtual network mapping task is ended;
and 3.3, judging whether the node entropies of the physical node i and the virtual node J are closest, if so, mapping the virtual node J to the physical node i, if J is J +1, judging whether J is less than J, if J is less than J, returning to the step 3.1, otherwise, successfully mapping the virtual node, and executing the step 4.
3. The node entropy-based virtual network mapping method according to claim 2,
in step 3.3, the method for judging whether the node entropies of the physical node i and the virtual node j are closest to each other is as follows: and calculating the absolute value of the node entropy difference between the virtual node j and each physical node, wherein if the absolute value of the node entropy difference between the physical node i and the virtual node j is the minimum in all the absolute values, the node entropy of the physical node i is closest to that of the virtual node j.
4. The method for mapping a virtual network based on node entropy as claimed in claim 1, wherein in step 4, when the target paths corresponding to multiple virtual links pass through the same physical link, a suitable target path is first found for the virtual link with a large bandwidth resource requirement, and then the virtual link with a small bandwidth requirement is considered.
5. The method for mapping a virtual network based on node entropy as claimed in claim 1, wherein in step 4, a shortest path algorithm is used to find a shortest target path between physical nodes bearing virtual nodes at two ends of each virtual link in the physical network for each virtual link, so as to perform virtual link mapping.
6. The method for mapping a virtual network based on node entropy as claimed in claim 1, wherein, assuming that there are L virtual links requested by a virtual network, step 4 specifically includes:
step 4.1, sequencing all virtual links from large to small according to bandwidth requirements, and putting results into Virtuallnklist;
step 4.2, deleting the physical link which does not meet the bandwidth requirement of the virtual link l for the virtual link l in the Virtuallnklist to obtain a physical link set which meets the bandwidth requirement of the virtual link l; if the set is empty, executing step 4.3, otherwise executing step 4.4;
step 4.3, the virtual link mapping fails, and the virtual network mapping task is finished;
and 4.4, mapping the virtual link L to the shortest physical link in the physical link set, enabling L to be L +1, judging whether L is less than L, if L is less than L, returning to the step 4.2, otherwise, successfully mapping the virtual link, and ending the virtual network mapping task.
CN201910464796.9A 2019-05-30 2019-05-30 Virtual network mapping method based on node entropy Active CN110213149B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910464796.9A CN110213149B (en) 2019-05-30 2019-05-30 Virtual network mapping method based on node entropy

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910464796.9A CN110213149B (en) 2019-05-30 2019-05-30 Virtual network mapping method based on node entropy

Publications (2)

Publication Number Publication Date
CN110213149A CN110213149A (en) 2019-09-06
CN110213149B true CN110213149B (en) 2021-02-26

Family

ID=67789759

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910464796.9A Active CN110213149B (en) 2019-05-30 2019-05-30 Virtual network mapping method based on node entropy

Country Status (1)

Country Link
CN (1) CN110213149B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112242951B (en) * 2020-10-16 2022-03-15 中国联合网络通信集团有限公司 Virtual network mapping method and device

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104486194A (en) * 2014-12-12 2015-04-01 北京邮电大学 Control system and control method for virtual network with multiple reliability levels
CN104993941A (en) * 2015-05-14 2015-10-21 西安电子科技大学 Openflow-based network highly-fault-tolerant virtual network mapping algorithm
CN105978713A (en) * 2016-05-06 2016-09-28 西安电子科技大学 Elastic optical network based resource distribution method in virtual network mapping
CN107147530A (en) * 2017-05-24 2017-09-08 西安交通大学 A kind of virtual network method for reconfiguration based on resource conservation
CN108667657A (en) * 2018-04-28 2018-10-16 西安交通大学 A kind of mapping method of virtual network based on local feature information towards SDN
CN109495300A (en) * 2018-11-07 2019-03-19 西安交通大学 A kind of reliable SDN virtual network mapping algorithm

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104486194A (en) * 2014-12-12 2015-04-01 北京邮电大学 Control system and control method for virtual network with multiple reliability levels
CN104993941A (en) * 2015-05-14 2015-10-21 西安电子科技大学 Openflow-based network highly-fault-tolerant virtual network mapping algorithm
CN105978713A (en) * 2016-05-06 2016-09-28 西安电子科技大学 Elastic optical network based resource distribution method in virtual network mapping
CN107147530A (en) * 2017-05-24 2017-09-08 西安交通大学 A kind of virtual network method for reconfiguration based on resource conservation
CN108667657A (en) * 2018-04-28 2018-10-16 西安交通大学 A kind of mapping method of virtual network based on local feature information towards SDN
CN109495300A (en) * 2018-11-07 2019-03-19 西安交通大学 A kind of reliable SDN virtual network mapping algorithm

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
Instance expansion algorithm for micro-service with prediction;Min Zhu，Hua Qu;《IEEE》;20180322;全文 *
基于全局资源容量的虚拟网络嵌入算法;孙新丽;《计算技术与自动化》;20190528;全文 *
拉普拉斯噪声下基于绝对值累积的动态到达频谱感知;赵季红;《计算机工程》;20181213;全文 *

Also Published As

Publication number Publication date
CN110213149A (en) 2019-09-06

Similar Documents

Publication Publication Date Title
Liao et al. Density cluster based approach for controller placement problem in large-scale software defined networkings
CN107566194B (en) Method for realizing cross-domain virtual network mapping
CN112738820B (en) Dynamic deployment method and device of service function chain and computer equipment
CN107094115B (en) Ant colony optimization load balancing routing algorithm based on SDN
Quang et al. Multi-domain non-cooperative VNF-FG embedding: A deep reinforcement learning approach
CN108322333B (en) Virtual network function placement method based on genetic algorithm
CN105515987B (en) A kind of mapping method based on SDN framework Virtual optical-fiber networks
CN113784373B (en) Combined optimization method and system for time delay and frequency spectrum occupation in cloud edge cooperative network
CN103917958A (en) Distributed mapping function for large scale media clouds
CN109379230B (en) Service function chain deployment method based on breadth-first search
CN107196806B (en) Topological proximity matching virtual network mapping method based on sub-graph radiation
CN111538570A (en) VNF deployment method and device for energy conservation and QoS guarantee
CN111162865A (en) Virtual optical network mapping method for sensing fragments in space division multiplexing elastic optical network
CN113742046A (en) Flow grooming cloud-side computing network computing resource balanced scheduling method and system
CN111866623A (en) High-efficiency virtual optical network survivability mapping method for service reliability
CN114071582A (en) Service chain deployment method and device for cloud-edge collaborative Internet of things
CN111800352B (en) Service function chain deployment method and storage medium based on load balancing
CN113300861B (en) Network slice configuration method, device and storage medium
Khodaparas et al. A software-defined caching scheme for the Internet of Things
CN111181792A (en) SDN controller deployment method and device based on network topology and electronic equipment
Quang et al. Evolutionary actor-multi-critic model for VNF-FG embedding
CN107360031B (en) Virtual network mapping method based on optimized overhead-to-revenue ratio
CN110213149B (en) Virtual network mapping method based on node entropy
Nguyen et al. Efficient virtual network embedding with node ranking and intelligent link mapping
CN111324429A (en) Micro-service combination scheduling method based on multi-generation ancestry reference distance

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant